Marine pH value automatic monitoring method and system
By using dynamic calibration cycles and edge-cloud collaborative anomaly detection, the problems of sensor drift and network interruption in ocean pH monitoring have been solved, thus improving the stability and reliability of the ocean pH monitoring system, extending its endurance, and enhancing the timeliness of early warnings.
Patent Information
- Application Number
- CN202510923606.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-04
- Publication Date
- 2025-10-21
AI Technical Summary
Existing ocean pH monitoring technology suffers from sensor drift and accuracy degradation in marine environments. The fixed-period calibration strategy cannot adapt to rapid changes in pH, and lacks an effective edge-cloud collaborative anomaly detection mechanism, resulting in an imbalance between monitoring accuracy and energy consumption, and anomalies cannot be identified in a timely manner when the network is interrupted.
By constructing a gradient acceleration parameter acquisition mechanism and a multi-parameter anomaly scoring function, the sensor calibration cycle is dynamically adjusted. Combined with a lightweight edge intelligence algorithm and an edge-cloud collaborative architecture, predictive calibration and real-time anomaly detection are achieved, and the monitoring mode is adaptively adjusted to ensure data accuracy and energy consumption balance.
It significantly improves the stability and continuity of the marine pH monitoring system in complex marine environments, extends the endurance of monitoring nodes, ensures data accuracy and timely early warning during critical periods of change, and provides more reliable technical support for marine environmental protection.
Smart Images
Figure CN120820604A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of ocean monitoring, and in particular to a method and system for automatically monitoring ocean pH. Background Art
[0002] Ocean pH is a key indicator of the health of marine ecosystems, and its precise monitoring is of great significance for assessing ocean acidification trends and protecting marine biodiversity. Traditional ocean pH measurements mainly rely on electrochemical analysis methods, such as the glass electrode method, ion-selective field-effect transistors (ISFETs), and solid-state ion-selective electrodes. These electrochemical sensors determine pH by measuring the potential difference or conductivity change caused by the activity of hydrogen ions in the solution. However, seawater, as a complex electrolyte solution, has a high salinity environment that affects the electrode interface potential, resulting in measurement deviations. At the same time, marine microorganisms easily form biofilms on the electrode surface, changing the electrode response characteristics. In addition, electrochemical sensors face challenges such as reference electrode drift and calibration difficulties during long-term ocean deployment. These factors together restrict the accuracy of ocean pH monitoring.
[0003] With the development of the Internet of Things (IoT), marine environmental monitoring systems based on wireless sensor networks are gaining popularity. A variety of existing marine monitoring solutions have been developed, including underwater parameter transmission technology based on underwater acoustic communications, remote water quality monitoring systems utilizing low-power wide-area network technologies such as LoRa, marine chemical parameter monitoring devices incorporating electrochemical sensing principles, and seawater pH inversion methods using machine learning. These technologies have, to varying degrees, advanced marine monitoring. In particular, neural network-based autocalibration methods and the application of cloud-edge collaborative architectures have provided a technical foundation for automated ocean pH monitoring. However, these solutions are primarily designed for terrestrial environments or static water bodies, making them difficult to directly apply to complex and volatile marine environments. In particular, long-term stable operation in open waters remains a challenge.
[0004] An in-depth analysis of existing technologies reveals two key deficiencies in current ocean pH monitoring technology. First, electrochemical pH sensors inevitably experience sensor drift and decreased accuracy during long-term operation in marine environments. Existing technologies generally adopt a fixed-cycle calibration strategy that does not consider the actual rate of change of ocean pH and sensor drift characteristics. During critical periods when ocean pH changes rapidly or sensor drift intensifies, important monitoring data may be missed due to long calibration cycles. Furthermore, during periods of stable ocean pH, overcalibration wastes limited battery energy. Second, existing technologies lack an effective edge-cloud collaborative anomaly detection mechanism and rely too heavily on cloud processing capabilities. When the ocean network is unstable or communication is interrupted, they cannot rely on local lightweight algorithms to maintain pH anomaly detection and warning. Especially in severe sea conditions, network communication interruptions may prevent timely identification and response to ocean pH anomalies, resulting in missed environmental risk warning opportunities. These problems seriously restrict the stable operation and accurate warning of ocean pH monitoring systems in complex marine environments. There is an urgent need to develop intelligent monitoring methods that are more adaptable to ocean characteristics.
[0005] Currently, no effective solutions have been proposed for the problems in related technologies. Summary of the Invention
[0006] In response to the problems in the related art, the present invention proposes a method and system for automatic monitoring of ocean pH, which has the advantages of intelligent sensor calibration based on the characteristics of ocean pH changes and edge-cloud collaborative anomaly detection, thereby solving the problems in the existing technology of fixed-period calibration leading to sensor drift and energy consumption imbalance, as well as the failure of monitoring functions during network interruptions.
[0007] To this end, the specific technical solutions adopted in the present invention are as follows:
[0008] According to one aspect of the present invention, a method for automatically monitoring ocean pH is provided, the method comprising:
[0009] S1. Based on the real-time ocean pH data from different monitoring nodes, the gradient acceleration parameters are obtained, the ocean pH change levels are classified, and the calibration cycle is determined by the calibration cycle calculation algorithm to perform predictive calibration.
[0010] S2. Build a multi-parameter anomaly scoring function based on the calibrated monitoring nodes, hierarchically store ocean pH data, and adjust the monitoring mode according to the wireless network status to achieve real-time detection and early warning of ocean pH anomalies.
[0011] S3. Summarize the calibration data, abnormal score values and warning information of each monitoring node, form an ocean pH monitoring report and output it.
[0012] Furthermore, based on the real-time ocean pH data from different monitoring nodes, gradient acceleration parameters are obtained to classify the ocean pH change levels. The calibration period is determined by the calibration period calculation algorithm. Predictive calibration is performed, including:
[0013] S11. Based on the real-time ocean pH data from different monitoring nodes, obtain the gradient acceleration parameters and classify the ocean pH change level;
[0014] S12. Determine the calibration period using a calibration period calculation algorithm based on the ocean pH change level and the energy consumption urgency coefficient;
[0015] S13. Construct different calibration trigger instructions based on the gradient acceleration parameters, perform calibration judgment, and perform predictive calibration according to the judgment result.
[0016] Furthermore, based on the real-time ocean pH data from different monitoring nodes, the gradient acceleration parameters are obtained and the ocean pH change levels are divided into the following categories:
[0017] Based on real-time ocean pH data from adjacent monitoring nodes in a distributed wireless sensor network, the spatial gradient change rate and temporal gradient change rate are calculated. The spatial gradient change rate is determined by the ratio of the pH difference between adjacent nodes to the distance between nodes, and the temporal gradient change rate is determined by the ratio of the pH difference of a single node at consecutive time points to the time interval.
[0018] According to the spatial gradient change rate and the temporal gradient change rate, the gradient acceleration parameter is defined as the second-order derivative of the temporal gradient change rate, and the ocean pH change trend is divided into different levels based on the positive and negative and absolute value of the gradient acceleration parameter.
[0019] Furthermore, according to the ocean pH change level and the energy consumption urgency coefficient, the calibration period is determined by the calibration period calculation algorithm, including:
[0020] The current battery voltage, historical energy consumption rate and estimated remaining working time of each monitoring node are obtained; the energy consumption urgency coefficient is calculated according to the ratio of the battery voltage to the standard working voltage; the calibration period is obtained by using the calibration period calculation algorithm based on the ocean pH change level and the energy consumption urgency coefficient; wherein, obtaining the calibration period by using the calibration period calculation algorithm includes: when the pH change level is rapidly deteriorating and the energy consumption urgency coefficient is greater than the second preset threshold, the calibration period is set to the first proportional multiple of the basic period; when the energy consumption urgency coefficient is less than the first preset threshold, regardless of the pH change level, the calibration period is extended to the second proportional multiple of the basic period; in other cases, the calibration period is determined based on the weighted calculation of the pH change level and the energy consumption urgency coefficient.
[0021] Furthermore, different calibration trigger instructions are constructed based on the gradient acceleration parameters, and calibration judgment is performed. Predictive calibration is performed according to the judgment results, including: when the gradient acceleration parameters are positive and increasing for a preset number of consecutive time points, it is determined that the ocean pH has entered a rapid change period, and a predictive calibration trigger instruction is generated; when the variance of the gradient acceleration parameters exceeds the fluctuation threshold, it is determined that the ocean pH fluctuation is abnormal, and an emergency calibration trigger instruction is generated; for the predictive calibration trigger instruction, if the predictive calibration trigger time is earlier than the periodic calibration time, a calibration priority identifier and a full-parameter calibration configuration are generated; for the emergency calibration trigger instruction, if the energy consumption urgency coefficient is less than the first preset threshold when the emergency calibration trigger instruction is generated, a fast calibration identifier and a key parameter calibration configuration are generated; when there is a calibration priority identifier, a full-parameter predictive calibration is performed and the cycle count is reset; when there is a fast calibration identifier, compressed time calibration is performed only for the parameters in the key parameter list; after the calibration is performed, the calibration status record of the monitoring node is updated, and the calibration result is transmitted to the adjacent nodes to achieve regional calibration data synchronization.
[0022] Furthermore, a multi-parameter anomaly scoring function is constructed based on the calibrated monitoring nodes to hierarchically store ocean pH data. The monitoring mode is adjusted according to the wireless network status to achieve real-time detection and early warning of ocean pH anomalies.
[0023] S21. Construct a multi-parameter anomaly scoring function based on the lightweight edge intelligence algorithm pre-deployed on the calibrated monitoring node and generate an anomaly score value;
[0024] S22. Based on the abnormality score value, the ocean pH data is stored in a hierarchical manner and the wireless network transmission order is set according to the priority;
[0025] S23. Automatically adjust the monitoring mode according to the wireless network status to achieve abnormal ocean pH detection and early warning for monitoring nodes under various network conditions.
[0026] Furthermore, based on the lightweight edge intelligent algorithm pre-deployed on the calibrated monitoring node, a multi-parameter anomaly scoring function is constructed, and an anomaly scoring value is generated, including: using a sliding time window to collect ocean pH measurement data, and calculating the short-term change rate and long-term change trend of the calibrated monitoring node; wherein the short-term change rate is the ratio of the difference between two adjacent measurement values to the time interval, and the long-term change trend is the first-order linear regression slope of the past N measurement values; based on the short-term change rate, long-term change trend and the current pH absolute value, an anomaly scoring function is constructed, and an anomaly scoring value is generated; based on the anomaly scoring value, when the score exceeds the high-risk threshold, an emergency warning is triggered and the sampling frequency is increased; when the score is between the high-risk threshold and the medium-risk threshold, a concern warning is triggered; when the score is below the medium-risk threshold, the regular monitoring state is maintained.
[0027] Furthermore, an anomaly scoring function is constructed and an anomaly score value is generated, including: when the ocean pH is below the healthy threshold, the anomaly score value is proportional to the degree to which the ocean pH deviates from the healthy threshold; when the short-term change rate exceeds the safe change rate, the anomaly score value is increased by a first weighting coefficient multiplied by the excess; when the long-term change trend continuously decreases for more than a preset time period, the anomaly score value is increased by a second weighting coefficient multiplied by the time excess;
[0028] The expression of the anomaly scoring function is:
[0029] AS(t)=α×max(0,pHh-pH(t))+β×max(0,|ΔpH(t)|-ΔpHs)
[0030] +γ×max(0,T d -T s );
[0031] Where AS(t) is the anomaly scoring function; pHh is the healthy pH threshold; pH(t) is the ocean pH value at the current time t; ΔpH(t) is the short-term change rate, and ΔpHs is the safe change rate threshold; T d is the duration of the long-term downward trend, T s is the preset time threshold; α is the pH deviation weight coefficient; β is the change rate weight coefficient; γ is the trend persistence weight coefficient.
[0032] Furthermore, the monitoring mode is automatically adjusted according to the wireless network status to realize the abnormal detection and early warning of ocean pH of monitoring nodes under various network conditions, including: judging the current status of the wireless network by detecting signal strength, data packet transmission success rate and network delay; when the network is in good condition, adopting cloud collaboration mode to send ocean pH data to the cloud for processing and storage; when the network is restricted, adopting hybrid mode, increasing local processing tasks, and only sending high-priority data; when the network is interrupted, adopting fully autonomous mode to complete data collection, storage and analysis locally; when it is detected that the network signal strength drops below the good signal threshold or the data packet transmission success rate drops below the stable transmission threshold, the local computing resource allocation ratio and data storage capacity are increased; when it is detected that the network signal strength rises above the minimum signal threshold and the data packet transmission success rate rises above the acceptable transmission threshold, the ocean pH data is uploaded synchronously and the cloud and local tasks are reallocated.
[0033] According to another aspect of the present invention, there is also provided an automatic ocean pH monitoring system, the automatic ocean pH monitoring system comprising:
[0034] The gradient calibration module is used to obtain gradient acceleration parameters based on the real-time ocean pH data from different monitoring nodes, classify the ocean pH change level, determine the calibration period through the calibration period calculation algorithm, and perform predictive calibration;
[0035] A dynamic monitoring module is used to construct a multi-parameter anomaly scoring function based on calibrated monitoring nodes, hierarchically store ocean pH data, and adjust the monitoring mode according to the wireless network status to achieve real-time detection and early warning of ocean pH anomalies;
[0036] The data integration module is used to summarize the calibration data, abnormal score values and warning information of each monitoring node, form and output the ocean pH monitoring report.
[0037] The beneficial effects of the present invention are:
[0038] (1) The method for automatic monitoring of ocean pH proposed in the present invention realizes intelligent dynamic adjustment of the sensor calibration period by constructing a gradient acceleration parameter acquisition mechanism and an ocean pH change level classification system, combined with the energy consumption urgency coefficient, and effectively solves the problem that traditional fixed-period calibration cannot adapt to the characteristics of ocean pH changes; at the same time, through the multi-parameter anomaly scoring function and edge-cloud collaborative architecture design, cloud data synchronization processing when the network is normal and local lightweight anomaly detection when the network is interrupted are realized, which significantly improves the stability of the ocean pH monitoring system in complex marine environments.
[0039] (2) Through the predictive calibration strategy based on gradient acceleration parameters, the present invention can dynamically adjust the calibration trigger conditions according to the actual change rate of ocean pH and sensor drift characteristics, increase the calibration frequency during the period of rapid pH change to ensure monitoring accuracy, and extend the calibration cycle during the stable period to save battery energy, thereby achieving an optimal balance between sensor calibration accuracy and energy consumption. Compared with the traditional fixed-cycle calibration method, it can extend the monitoring node life time by more than 30% and ensure data accuracy during the critical change period.
[0040] (3) Through the local deployment of lightweight edge intelligent algorithms and a multi-mode monitoring switching mechanism, the present invention achieves an adaptive response to the network communication status. When the network is normal, cloud data synchronization and complex analysis are performed through the wireless network. When the network is interrupted, it automatically switches to the local anomaly detection mode to continue to maintain the monitoring and early warning functions, ensuring the continuity and reliability of marine pH monitoring, avoiding monitoring blind spots and early warning losses caused by network failures, and providing more reliable technical support for marine environmental protection. BRIEF DESCRIPTION OF THE DRAWINGS
[0041] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0042] Figure 1 is a flow chart of a method for automatically monitoring ocean pH according to an embodiment of the present invention;
[0043] Figure 2 This is a principle block diagram of an automatic ocean pH monitoring system according to an embodiment of the present invention. DETAILED DESCRIPTION
[0044] To further illustrate each embodiment, the present invention provides drawings, which are part of the disclosure of the present invention. They are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. By referring to these contents, ordinary technicians in this field should be able to understand other possible implementation methods and advantages of the present invention. The components in the figures are not drawn to scale, and similar component symbols are generally used to represent similar components.
[0045] According to an embodiment of the present invention, a method and system for automatically monitoring ocean pH are provided.
[0046] The present invention will now be further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to one embodiment of the present invention, a method for automatically monitoring ocean pH is provided, and the method for automatically monitoring ocean pH includes:
[0047] S1. Based on the real-time ocean pH data from different monitoring nodes, the gradient acceleration parameters are obtained, the ocean pH change levels are classified, and the calibration cycle is determined by the calibration cycle calculation algorithm to perform predictive calibration.
[0048] S2. Build a multi-parameter anomaly scoring function based on the calibrated monitoring nodes, hierarchically store ocean pH data, and adjust the monitoring mode according to the wireless network status to achieve real-time detection and early warning of ocean pH anomalies.
[0049] S3. Summarize the calibration data, abnormal score values and warning information of each monitoring node, form an ocean pH monitoring report and output it.
[0050] Specifically, the method for automatically monitoring ocean pH is applied to a distributed wireless sensor network consisting of multiple monitoring nodes, each of which includes an electrochemical pH sensor (pH sensor), a temperature sensor, a computing and processing unit, a wireless communication unit, and an energy management unit. The electrochemical pH sensor uses a glass electrode method to measure the pH of seawater, which has the characteristics of high measurement accuracy and short response time. The temperature sensor is used to collect ambient temperature data and provide temperature compensation for pH measurement. The computing and processing unit uses a low-power microcontroller (such as the ARM Cortex-M series) with edge computing capabilities to deploy lightweight edge intelligent algorithms, process sensor data, perform calibration, and detect anomalies. The wireless communication unit uses low-power wide area network technology (such as LoRa or NB-IoT) to achieve data transmission between monitoring nodes and communication with the cloud. The energy management unit contains a solar charging device and a lithium battery pack, which extends the node working time through an intelligent power consumption control algorithm. Together, these components constitute the hardware foundation of the method for automatically monitoring ocean pH.
[0051] In one embodiment, based on real-time ocean pH data from different monitoring nodes, gradient acceleration parameters are obtained, ocean pH change levels are classified, and a calibration period is determined using a calibration period calculation algorithm. Performing predictive calibration includes:
[0052] S11. Based on the real-time ocean pH data from different monitoring nodes, obtain the gradient acceleration parameters and classify the ocean pH change level;
[0053] S12. Determine the calibration period using a calibration period calculation algorithm based on the ocean pH change level and the energy consumption urgency coefficient;
[0054] S13. Construct different calibration trigger instructions based on the gradient acceleration parameters, perform calibration judgment, and perform predictive calibration according to the judgment result.
[0055] In one embodiment, based on the real-time ocean pH data from different monitoring nodes, gradient acceleration parameters are obtained and the ocean pH change levels are classified as follows:
[0056] S111. Calculate the spatial gradient change rate and the temporal gradient change rate based on the real-time ocean pH data from adjacent monitoring nodes in the distributed wireless sensor network.
[0057] The spatial gradient change rate is determined by the ratio of the pH difference between adjacent nodes to the distance between nodes, and the temporal gradient change rate is determined by the ratio of the pH difference of a single node at consecutive time points to the time interval.
[0058] S112. Based on the spatial gradient change rate and the temporal gradient change rate, define the gradient acceleration parameter as the second-order derivative of the temporal gradient change rate, and classify the ocean pH change trend into different levels based on the positive and negative sign and absolute value of the gradient acceleration parameter;
[0059] Specifically, the levels of ocean pH change include rapid deterioration, slow deterioration, stable state, slow improvement and rapid improvement.
[0060] Specifically, in step S11, real-time ocean pH data from multiple monitoring nodes is first collected from a distributed wireless sensor network and transmitted to a data processing center via wireless communication units. Spatial gradients are calculated using data from adjacent monitoring nodes, and the spatial distribution of ocean pH is identified by comparing pH differences between nodes at different locations. Furthermore, temporal gradients are calculated using data from consecutive time points at a single node to analyze the temporal variation trend of ocean pH. Based on this gradient information, a gradient acceleration parameter (the second-order derivative of the temporal gradient rate of change) is further calculated. This parameter effectively reflects the acceleration or deceleration trend of ocean pH changes. According to the numerical characteristics of the gradient acceleration parameter, the changes in ocean pH are divided into five levels: rapid deterioration (gradient acceleration is negative and the absolute value is greater than the upper threshold), slow deterioration (gradient acceleration is negative and the absolute value is between the upper and lower thresholds), stable state (gradient acceleration is close to zero), slow improvement (gradient acceleration is positive and the value is between the upper and lower thresholds) and rapid improvement (gradient acceleration is positive and the value is greater than the upper threshold). This classification method enables the accurate identification of the change pattern of ocean pH and provides a basis for the formulation of subsequent calibration strategies.
[0061] In one embodiment, determining the calibration period using a calibration period calculation algorithm based on the ocean pH change level and the energy consumption urgency coefficient includes:
[0062] S121. Obtain the current battery voltage, historical energy consumption rate, and estimated remaining operating time of each monitoring node; calculate the energy consumption urgency coefficient based on the ratio of the battery voltage to the standard operating voltage;
[0063] Specifically, when the energy consumption emergency coefficient is less than a first preset threshold, it is defined as an energy consumption emergency state.
[0064] S122. Based on the ocean pH change level and the energy consumption urgency coefficient, a calibration period calculation algorithm is used to obtain a calibration period;
[0065] Using the calibration period calculation algorithm, obtaining the calibration period includes:
[0066] When the pH change level deteriorates sharply and the energy consumption urgency coefficient is greater than the second preset threshold, the calibration period is set to the first proportional multiple of the basic period; when the energy consumption urgency coefficient is less than the first preset threshold, the calibration period is extended to the second proportional multiple of the basic period regardless of the pH change level; in other cases, the calibration period is determined based on the weighted calculation of the pH change level and the energy consumption urgency coefficient.
[0067] Specifically, in step S12, the calibration period is dynamically adjusted by comprehensively considering the change level of ocean pH and the energy status of the monitoring node. First, the current battery voltage, historical energy consumption rate, and estimated remaining working time of each monitoring node are obtained through the energy management unit; the ratio of the current battery voltage to the standard working voltage is defined as the energy consumption urgency coefficient, which reflects the urgency of the energy status of the monitoring node. When the energy consumption urgency coefficient is less than the first preset threshold (such as 0.3), it is determined that the node is in an energy consumption emergency state and energy-saving measures need to be taken. Based on the change level of ocean pH and the energy consumption urgency coefficient, the calibration period calculation algorithm is applied to determine the optimal calibration period. The specific algorithm logic is as follows: when the pH change level is rapidly deteriorating and the energy consumption urgency coefficient is greater than a second preset threshold (such as 0.7), the calibration period is set to a first proportional multiple (such as 0.5 times, i.e., 12 hours) of the basic period (such as 24 hours) to increase the calibration frequency to cope with the rapidly changing marine environment; when the energy consumption urgency coefficient is less than the first preset threshold, regardless of the pH change level, the calibration period is extended to a second proportional multiple (such as 2 times, i.e., 48 hours) of the basic period to save energy; in other cases, the calibration period is determined based on a weighted calculation of the pH change level and the energy consumption urgency coefficient, with the weighted calculation formula being: calibration period = basic period × (1-0.3 × change level coefficient - 0.2 × (1-energy consumption urgency coefficient)), where the change level coefficient corresponds to 1.0, 0.75, 0.5, 0.25, and 0 from rapidly deteriorating to rapidly improving, respectively. By dynamically adjusting the calibration period, the present invention can both meet the accuracy requirements of marine environmental monitoring and optimize energy utilization efficiency.
[0068] In one embodiment, constructing different calibration trigger instructions based on the gradient acceleration parameters, performing calibration judgment, and performing predictive calibration according to the judgment result includes:
[0069] S131. When the gradient acceleration parameter is positive and increasing for a predetermined number of consecutive time points, it is determined that the ocean pH has entered a period of rapid change, and a predictive calibration trigger instruction is generated. When the variance of the gradient acceleration parameter exceeds a fluctuation threshold, it is determined that the ocean pH fluctuation is abnormal, and an emergency calibration trigger instruction is generated.
[0070] S132. For a predictive calibration trigger instruction, if the predictive calibration trigger time is earlier than the periodic calibration time, a calibration priority identifier and a full-parameter calibration configuration are generated; for an emergency calibration trigger instruction, if the energy consumption urgency coefficient is less than a first preset threshold when the emergency calibration trigger instruction is generated, a quick calibration identifier and a key parameter calibration configuration are generated;
[0071] S133. When a calibration priority flag is present, perform full-parameter predictive calibration and reset the cycle count; when a fast calibration flag is present, perform compressed time calibration only on the parameters in the key parameter list; after performing the calibration, update the calibration status record of the monitoring node, and transmit the calibration result to the adjacent nodes to achieve regional calibration data synchronization.
[0072] Specifically, in step S13, an intelligent calibration trigger mechanism is constructed based on the gradient acceleration parameter to achieve predictive calibration. The changing trend of the gradient acceleration parameter is continuously monitored. When the parameter is positive and the value increases for a preset number of time points (set to 3-5 in this embodiment), it is determined that the ocean pH has entered a rapid change period. At this time, a predictive calibration trigger instruction is generated, and calibration is performed in advance to ensure measurement accuracy. When the variance of the gradient acceleration parameter exceeds the fluctuation threshold (set to 2 times the standard deviation in this embodiment), it is determined that the ocean pH fluctuation is abnormal, and an emergency calibration trigger instruction is generated. For the predictive calibration trigger instruction, it is further determined whether its trigger time is earlier than the periodic calibration time. If so, a calibration priority flag and a full parameter calibration configuration are generated to indicate that a full calibration should be performed in advance. For the emergency calibration trigger instruction, the current energy consumption urgency coefficient is checked. If it is less than the first preset threshold, a quick calibration flag and a key parameter calibration configuration are generated to ensure the measurement accuracy of key parameters while saving energy. During the actual calibration process, when a calibration priority indicator is present, a full-parameter predictive calibration is performed and the cycle count is reset. When a quick calibration indicator is present, a compressed time calibration is performed only on the parameters in the key parameter list (pH value, temperature). Compressed time calibration improves efficiency by reducing the number of calibration points and shortening the stabilization wait time. After calibration is completed, the calibration status record of the monitoring node is updated, including information such as calibration time, calibration parameters, calibration results, and calibration deviation. These calibration results are then transmitted to adjacent monitoring nodes via wireless communication units to achieve regional calibration data synchronization.
[0073] In one embodiment, a multi-parameter anomaly scoring function is constructed based on calibrated monitoring nodes, ocean pH data is stored hierarchically, and the monitoring mode is adjusted according to the wireless network status to achieve real-time detection and early warning of ocean pH anomalies. The following steps are included:
[0074] S21. Construct a multi-parameter anomaly scoring function based on the lightweight edge intelligence algorithm pre-deployed on the calibrated monitoring node and generate an anomaly score value;
[0075] S22. Based on the abnormality score value, the ocean pH data is stored in a hierarchical manner and the wireless network transmission order is set according to the priority;
[0076] S23. Automatically adjust the monitoring mode according to the wireless network status to achieve abnormal ocean pH detection and early warning for monitoring nodes under various network conditions.
[0077] In one embodiment, a multi-parameter anomaly scoring function is constructed based on a lightweight edge intelligence algorithm pre-deployed on a calibrated monitoring node, and the anomaly score value is generated, including:
[0078] S211. Collect ocean pH measurement data using a sliding time window and calculate the short-term rate of change and long-term trend of the calibrated monitoring nodes. The short-term rate of change is the ratio of the difference between two consecutive measurements to the time interval, and the long-term trend is the slope of the first-order linear regression of the past N measurements.
[0079] S212. Construct an anomaly scoring function based on the short-term change rate, the long-term change trend, and the current absolute value of pH, and generate an anomaly scoring value;
[0080] S213. Based on the abnormal score value, when the score exceeds the high-risk threshold, an emergency warning is triggered and the sampling frequency is increased; when the score is between the high-risk threshold and the medium-risk threshold, a concern warning is triggered; when the score is below the medium-risk threshold, the regular monitoring status is maintained.
[0081] In one embodiment, constructing an anomaly scoring function and generating an anomaly scoring value includes:
[0082] When ocean pH is below the healthy threshold, the anomaly score is proportional to the extent to which the ocean pH deviates from the healthy threshold;
[0083] When the short-term change rate exceeds the safe change rate, the abnormal score value increases by the first weighting coefficient multiplied by the excess;
[0084] When the long-term trend continues to decline for more than a preset period of time, the abnormal score value is increased by a second weighting coefficient multiplied by the time excess;
[0085] The expression of the anomaly scoring function is:
[0086] AS(t)=α×max(0,pHh-pH(t))+β×max(0,|ΔpH(t)|-ΔpHs)
[0087] +γ×max(0,T d -T s );
[0088] Where, AS(t) is the anomaly scoring function; pHh is the healthy pH threshold; pH(t) is the ocean pH value at the current time t; ΔpH(t) is the short-term change rate, and ΔpHs is the safety change rate threshold; T d is the duration of the long-term downward trend, T s is the preset time threshold; α is the pH deviation weight coefficient, which is inversely proportional to the gradient acceleration parameter defined in S1; β is the change rate weight coefficient, which is associated with the ocean pH change level calculated in S1; γ is the trend duration weight coefficient, which decreases as the energy consumption emergency coefficient increases.
[0089] It should be noted that max(0, pHh - pH(t)) is used to evaluate the degree to which the current ocean pH value deviates from the healthy range. The specific explanation is as follows:
[0090] pHh is the healthy pH threshold, and pH(t) is the ocean pH value at the current time t;
[0091] pHh - pH(t) calculates the difference between the current ocean pH value and the healthy threshold;
[0092] The max(0, pHh - pH(t)) function ensures that a positive value is generated only when pH(t) is lower than the healthy threshold pHh;
[0093] The function designed in this invention has the following functions:
[0094] When pH(t) < pHh (the ocean pH value is lower than the healthy threshold), the result is pHh - pH(t), that is, the degree of deviation;
[0095] When pH(t) ≥ pHh (the ocean pH value is within the healthy range), the result is 0, and the anomaly score is not increased;
[0096] This invention adopts this "penalty function" form, and the score is increased only when the parameter exceeds the safe range, which is very suitable for describing the mathematical model of ocean pH anomaly detection. Similarly, the other two max functions in the formula follow a similar principle:
[0097] max(0, |ΔpH(t)| - ΔpHs) is included only when the change rate exceeds the safe threshold;
[0098] max(0, T d - T s ) is included only when the trend duration exceeds the preset threshold.
[0099] Specifically, in step S21, the ocean pH data provided by the electrochemical pH sensor calibrated in step S1 is used to achieve near-real-time anomaly detection through a lightweight edge intelligence algorithm pre-deployed on the computing processing unit of the monitoring node. This edge intelligence algorithm uses a simplified machine learning model with low resource consumption and high computational efficiency, and is particularly suitable for running on monitoring nodes with limited computing power and storage space. First, continuous ocean pH measurement data is collected using a sliding time window technology. In this embodiment, the window size is 24 hours and the sliding step is 1 hour. Based on the collected ocean pH data, two key indicators are calculated: the short-term change rate and the long-term change trend. The short-term change rate reflects the rapid change of ocean pH between two adjacent measurements. The calculation method is the ratio of the difference between the two adjacent measurement values to the time interval; the long-term change trend is obtained by performing a first-order linear regression analysis on the past N measurements (such as 48-72 times). The slope of the regression line represents the long-term change trend. These two indicators are combined with the current absolute value of pH to construct a multi-parameter anomaly scoring function, which comprehensively considers the degree of deviation, rate of change and continuous trend of ocean pH, and can comprehensively evaluate the state of the marine environment. Based on the calculated anomaly score value, a three-level early warning mechanism is implemented: when the score exceeds the high-risk threshold (such as 80 points), an emergency warning is triggered and the sampling frequency is automatically increased from the standard once an hour to once every 15 minutes to capture rapidly changing environmental conditions; when the score is between the high-risk threshold and the medium-risk threshold (such as 50 points), a concern warning is triggered to remind relevant personnel to pay close attention to changes in ocean pH; when the score is lower than the medium-risk threshold, the routine monitoring state is maintained and the standard sampling frequency is maintained. The present invention significantly improves the timeliness of early warning for ocean pH monitoring through this multi-parameter-based anomaly scoring mechanism.
[0100] Specifically, in step S22, based on the anomaly score generated in S21, an intelligent hierarchical storage and priority transmission mechanism for ocean pH data is implemented. First, the data is divided into three storage levels based on the anomaly score: high-priority data (data with anomaly scores exceeding the high-risk threshold), medium-priority data (data with anomaly scores between the high-risk and medium-risk thresholds), and low-priority data (data with anomaly scores below the medium-risk threshold). For high-priority data, a redundant storage strategy is adopted, storing the data in both the local memory of the monitoring node and the backup area of adjacent nodes. For medium-priority data, only the complete data is stored in the local memory. For low-priority data, a compressed storage method is used, storing only data and statistical summary information at key time points to save storage space. In terms of data transmission, different network transmission strategies are set according to priority: high-priority data adopts real-time transmission mode, attempting to immediately transmit data to the cloud server regardless of network conditions; medium-priority data adopts quasi-real-time transmission mode, transmitting when the network load is low; and low-priority data adopts batch transmission mode, packaging data from multiple time points and transmitting them all at once when network conditions are good, to improve network utilization efficiency. When it detects a decrease in network bandwidth, it automatically reduces the transmission frequency of low-priority data to ensure the transmission quality of high-priority data. When the network condition recovers, it automatically clears the backlog of data queues and transmits high-priority data first, followed by medium-priority and low-priority data.
[0101] In one embodiment, automatically adjusting the monitoring mode according to the wireless network status to achieve abnormal ocean pH detection and early warning for monitoring nodes under various network conditions includes:
[0102] S231, determining the current state of the wireless network by detecting signal strength, data packet transmission success rate, and network delay;
[0103] Specifically, when the signal strength is higher than the good signal threshold, the data packet transmission success rate is higher than the stable transmission threshold, and the network delay is lower than the fast response threshold, the network is judged to be in a good state; when the signal strength is lower than the minimum signal threshold, the data packet transmission success rate is lower than the acceptable transmission threshold, or the network delay is higher than the allowable delay threshold, the network is judged to be in an interrupted state; otherwise, the network is judged to be in a restricted state.
[0104] S232. When the network is good, a cloud collaboration mode is used to send ocean pH data to the cloud for processing and storage. When the network is limited, a hybrid mode is used to increase local processing tasks and only send high-priority data. When the network is disconnected, a fully autonomous mode is used to complete data collection, storage and analysis locally.
[0105] S233. When the network signal strength is detected to drop below the good signal threshold or the data packet transmission success rate drops below the stable transmission threshold, increase the local computing resource allocation ratio and data storage capacity; when the network signal strength is detected to rise above the minimum signal threshold and the data packet transmission success rate rises above the acceptable transmission threshold, perform synchronous upload of ocean pH data and reallocate cloud and local tasks.
[0106] Specifically, in step S23, the monitoring mode is intelligently adjusted based on the wireless network status to ensure continuous detection and early warning of ocean pH anomalies under various network conditions. First, the wireless network status is evaluated using three key indicators: signal strength, packet transmission success rate, and network latency. Based on these indicators, the network status is divided into three categories: good network status, restricted network status, and disconnected network status. When the signal strength is above the good signal threshold (e.g., -70dBm), the packet transmission success rate is above the stable transmission threshold (e.g., 95%), and the network latency is below the fast response threshold (e.g., 200ms), the network is considered to be in a good state. When the signal strength is below the minimum signal threshold (e.g., -100dBm), the packet transmission success rate is below the acceptable transmission threshold (e.g., 60%), or the network latency is above the tolerable latency threshold (e.g., 1000ms), the network is considered to be in a disconnected state. All situations other than the above two are considered to be restricted network status. Under different network conditions, corresponding monitoring modes are adopted: when the network is good, a cloud collaboration mode is adopted, and the collected ocean pH data is sent to the cloud server in real time for processing and storage; when the network is restricted, a hybrid mode is adopted, increasing the proportion of local edge computing tasks, sending only high-priority data (such as abnormal event data and warning information) to the cloud, and temporarily storing the rest of the data locally; when the network is interrupted, it switches to a fully autonomous mode, and all data collection, storage, and analysis tasks are completed locally on the monitoring node, automatically increasing local storage space allocation and reducing the execution frequency of non-critical tasks to extend working time. When the network signal strength is detected to drop below the good signal threshold or the data packet transmission success rate drops below the stable transmission threshold, the local computing resource allocation ratio and data storage capacity are proactively increased to cope with possible network deterioration; when the network signal strength is detected to rise above the minimum signal threshold and the data packet transmission success rate rises above the acceptable transmission threshold, the ocean pH data synchronization upload operation is executed, uploading the locally stored historical data to the cloud, and reallocating cloud and local tasks based on the latest network conditions. The present invention improves the reliability of automatic monitoring of ocean pH in complex and changeable ocean environments through the adaptive monitoring mode switching mechanism based on network status.
[0107] Specifically, in step S3, three types of data are collected from each monitoring node in the distributed wireless sensor network: ocean pH measurement data from calibrated electrochemical pH sensors, anomaly scores generated using a multi-parameter anomaly scoring function, and various levels of warning information. This data is integrated to generate a monitoring report, which is then transmitted to a cloud-based management terminal via wireless communication units, providing data support for marine environmental monitoring.
[0108] like Figure 2 According to another embodiment of the present invention, there is also provided an automatic ocean pH monitoring system, the automatic ocean pH monitoring system comprising:
[0109] Gradient calibration module 1 is used to obtain gradient acceleration parameters based on the real-time ocean pH data of different monitoring nodes, classify the ocean pH change level, determine the calibration period through the calibration period calculation algorithm, and perform predictive calibration;
[0110] Dynamic Monitoring Module 2 is used to construct a multi-parameter anomaly scoring function based on calibrated monitoring nodes, hierarchically store ocean pH data, and adjust the monitoring mode according to the wireless network status to achieve real-time detection and early warning of ocean pH anomalies;
[0111] The data integration module 3 is used to summarize the calibration data, abnormal score values and warning information of each monitoring node, form an ocean pH monitoring report and output it.
[0112] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.
Claims
1. A method for automatically monitoring ocean pH, characterized in that: include: S1. Based on the real-time ocean pH data from different monitoring nodes, the gradient acceleration parameters are obtained, the ocean pH change levels are classified, and the calibration cycle is determined by the calibration cycle calculation algorithm to perform predictive calibration. S2. Build a multi-parameter anomaly scoring function based on the calibrated monitoring nodes, hierarchically store ocean pH data, and adjust the monitoring mode according to the wireless network status to achieve real-time detection and early warning of ocean pH anomalies. S3. Summarize the calibration data, abnormal score values and warning information of each monitoring node, form an ocean pH monitoring report and output it.
2. The method for automatically monitoring ocean pH according to claim 1, characterized in that: The method of obtaining gradient acceleration parameters based on the real-time ocean pH data from different monitoring nodes, classifying the ocean pH change levels, and determining the calibration period through a calibration period calculation algorithm to perform predictive calibration includes: S11. Based on the real-time ocean pH data from different monitoring nodes, obtain the gradient acceleration parameters and classify the ocean pH change levels; S12. Determine the calibration period using a calibration period calculation algorithm based on the ocean pH change level and the energy consumption urgency coefficient; S13. Construct different calibration trigger instructions based on the gradient acceleration parameters, perform calibration judgment, and perform predictive calibration according to the judgment result.
3. The method for automatically monitoring ocean pH according to claim 2, wherein: The method of obtaining gradient acceleration parameters based on real-time ocean pH data from different monitoring nodes and classifying ocean pH change levels includes: Based on real-time ocean pH data from adjacent monitoring nodes in a distributed wireless sensor network, the spatial gradient change rate and the temporal gradient change rate are calculated; wherein the spatial gradient change rate is determined by the ratio of the pH difference between adjacent nodes to the distance between nodes, and the temporal gradient change rate is determined by the ratio of the pH difference of a single node at consecutive time points to the time interval; According to the spatial gradient change rate and the temporal gradient change rate, the gradient acceleration parameter is defined as the second-order derivative of the temporal gradient change rate, and the ocean pH change trend is divided into different levels based on the positive and negative and absolute value of the gradient acceleration parameter.
4. The method for automatically monitoring ocean pH according to claim 2, wherein: Determining the calibration period by a calibration period calculation algorithm based on the ocean pH change level and the energy consumption urgency coefficient includes: Obtain the current battery voltage, historical energy consumption rate, and estimated remaining operating time of each monitoring node; calculate the energy consumption urgency coefficient based on the ratio of the battery voltage to the standard operating voltage; Based on the ocean pH change level and energy consumption urgency coefficient, the calibration period is obtained using the calibration period calculation algorithm; Among them, the use of the calibration period calculation algorithm to obtain the calibration period includes: when the pH change level is rapidly deteriorating and the energy consumption urgency coefficient is greater than the second preset threshold, the calibration period is set to the first proportional multiple of the basic period; when the energy consumption urgency coefficient is less than the first preset threshold, regardless of the pH change level, the calibration period is extended to the second proportional multiple of the basic period; in other cases, the calibration period is determined based on the weighted calculation of the pH change level and the energy consumption urgency coefficient.
5. The method for automatically monitoring ocean pH according to claim 2, wherein: The constructing of different calibration trigger instructions based on the gradient acceleration parameters, performing calibration judgment, and performing predictive calibration according to the judgment result includes: When the gradient acceleration parameter is positive and increasing for a preset number of consecutive time points, it is determined that the ocean pH has entered a period of rapid change and a predictive calibration trigger instruction is generated. When the variance of the gradient acceleration parameter exceeds the fluctuation threshold, it is determined that the ocean pH fluctuation is abnormal and an emergency calibration trigger instruction is generated. For a predictive calibration trigger instruction, if the predictive calibration trigger time is earlier than the periodic calibration time, a calibration priority identifier and a full-parameter calibration configuration are generated; for an emergency calibration trigger instruction, if the energy consumption urgency coefficient is less than a first preset threshold when the emergency calibration trigger instruction is generated, a quick calibration identifier and a key parameter calibration configuration are generated; When a calibration priority flag is present, full-parameter predictive calibration is performed and the cycle count is reset; when a fast calibration flag is present, compressed time calibration is performed only on the parameters in the key parameter list; after calibration is performed, the calibration status record of the monitoring node is updated, and the calibration results are transmitted to adjacent nodes to achieve regional calibration data synchronization.
6. The method for automatically monitoring ocean pH according to claim 1, characterized in that: The multi-parameter anomaly scoring function is constructed based on the calibrated monitoring nodes, the ocean pH data is stored hierarchically, and the monitoring mode is adjusted according to the wireless network status to achieve real-time detection and early warning of ocean pH anomalies. The method includes: S21. Construct a multi-parameter anomaly scoring function based on the lightweight edge intelligence algorithm pre-deployed on the calibrated monitoring node and generate an anomaly score value; S22. Based on the abnormality score value, the ocean pH data is stored in a hierarchical manner and the wireless network transmission order is set according to the priority; S23. Automatically adjust the monitoring mode according to the wireless network status to achieve abnormal ocean pH detection and early warning for monitoring nodes under various network conditions.
7. The method for automatically monitoring ocean pH according to claim 6, characterized in that: The method of constructing a multi-parameter anomaly scoring function based on the lightweight edge intelligent algorithm pre-deployed on the calibrated monitoring node and generating an anomaly score value includes: Using a sliding time window to collect ocean pH measurement data, the short-term rate of change and long-term trend of the calibrated monitoring nodes are calculated; the short-term rate of change is the ratio of the difference between two consecutive measurements to the time interval, and the long-term trend is the first-order linear regression slope of the past N measurements; Construct an anomaly scoring function based on the short-term change rate, long-term change trend and the current pH absolute value, and generate an anomaly score value; Based on the abnormal score value, when the score exceeds the high-risk threshold, an emergency warning is triggered and the sampling frequency is increased; when the score is between the high-risk threshold and the medium-risk threshold, a concern warning is triggered; when the score is below the medium-risk threshold, the regular monitoring status is maintained.
8. The method for automatically monitoring ocean pH according to claim 7, characterized in that: The process of constructing an anomaly scoring function and generating an anomaly scoring value includes: When the ocean pH falls below the healthy threshold, the anomaly score is proportional to the extent to which the ocean pH deviates from the healthy threshold. When the short-term rate of change exceeds the safe rate of change, the anomaly score is increased by the first weighting coefficient multiplied by the excess. When the long-term trend of change continues to decline for more than a preset time period, the anomaly score is increased by the second weighting coefficient multiplied by the time excess. The expression of the abnormality scoring function is: Where AS(t) is the anomaly scoring function; pHh is the healthy pH threshold; pH(t) is the ocean pH value at the current time t; ΔpH(t) is the short-term change rate, and ΔpHs is the safe change rate threshold; T d is the duration of the long-term downward trend, T s is the preset time threshold; α is the pH deviation weight coefficient; β is the change rate weight coefficient; γ is the trend persistence weight coefficient.
9. The method for automatically monitoring ocean pH according to claim 6, characterized in that: The automatic adjustment of the monitoring mode according to the wireless network status to achieve abnormal ocean pH detection and early warning for monitoring nodes under various network conditions includes: Determine the current status of the wireless network by detecting signal strength, data packet transmission success rate and network latency; When the network is good, the ocean pH data is sent to the cloud for processing and storage. When the network is limited, local processing tasks are added and only high-priority data is sent. When the network is disconnected, data collection, storage and analysis are completed locally. When the network signal strength is detected to drop below the good signal threshold or the data packet transmission success rate drops below the stable transmission threshold, the local computing resource allocation ratio and data storage capacity are increased; when the network signal strength is detected to rise above the minimum signal threshold and the data packet transmission success rate rises above the acceptable transmission threshold, the ocean pH data is uploaded synchronously and the cloud and local tasks are reallocated.
10. An automatic ocean pH monitoring system for implementing the automatic ocean pH monitoring method according to any one of claims 1 to 9, characterized in that: The system includes: The gradient calibration module is used to obtain gradient acceleration parameters based on the real-time ocean pH data from different monitoring nodes, classify the ocean pH change level, determine the calibration period through the calibration period calculation algorithm, and perform predictive calibration; A dynamic monitoring module is used to construct a multi-parameter anomaly scoring function based on calibrated monitoring nodes, hierarchically store ocean pH data, and adjust the monitoring mode according to the wireless network status to achieve real-time detection and early warning of ocean pH anomalies; The data integration module is used to summarize the calibration data, abnormal score values and warning information of each monitoring node, form and output the ocean pH monitoring report.